Wiki

Clone wiki

FE2AT / FE2AT_Usage_details

#Usage details of FE2AT (a.k.a How to use FE2AT)# FE2AT is a command line driven program that must be compiled from source. The current stable version of FE2AT is V3.0.

The execution of FE2AT requires a parameter file, the details of which are also provided below.

##1. Compilation (in the source directory)##

  • with the Intel Fortran Compiler (verified with version 12.1.2):

    ifort -vec_report0 -extend_source ./FE2AT_Vx.x.f90 -o FE2AT_Vx.x

  • with the GNU Fortran Compiler (verified with version 4.3.4):

    gfortran ./FE2AT_Vx.x.f90 -o ./FE2AT_Vx.x

    PLEASE NOTE: It is not verified that FE2AT works properly when compiled with the latest GNU Fortran Compiler!

##2. General usage##

./FE2AT_Vx.x -p [parameter file]

##3. Parameter file

| PARAMETER NAME | PARAMETER TYPE | DESCRIPTION | |-------------------|-------------------|------------- | | elementfile | [filename] | Definition of elements (either hexahedral, tetrahedral or quad elements - no mixture of them)
FORMAT DESCRIPTION (ABAQUS): - 1st column: element number, - following columns (4 for 2D quad and 3D tetrahedral, 8 for 3D hexahedral elements): numbers of the nodes | | nodefile | [filename] | Initial (undeformed) coordinates of the nodes FORMAT DESCRIPTION (ABAQUS): - 1st column: node number - following columns (2 for 2D, 3 for 3D): nodal coordinates (x,y[,z]) | | dispfile | [filename] | Nodal displacements due to the deformation FORMAT DESCRIPTION (ABAQUS): - 1st column: node number - following columns (2 for 2D, 3 for 3D): nodal displacements (ux,uy[,uz]) | | coordfile | [filename] | Atomic confguration file where the nodal displacements should be mapped onto; The atomic coordinates should correspond to the length-scale and dimensions of the FE mesh, i.e., atoms which lie not within the FE mesh will not be printed in the output configuration. Those atoms are finally written in the file FE2ATmap_output_notallocated.xyz. FORMAT DESCRIPTION (IMD): - 1st-3rd columns: number, type and mass of the atom - 4th-6th columns: atomic positions (x,y,z) - box dimensions are read form the imd-formatted header | | outfile | [filename] | Name of the deformed atomic output configuration FORMAT DESCRIPTION (IMD): - 1st-3rd columns: as in input file - 4th-6th columns: new atomic positions (x,y,z) after deformation - 7th column:,number of the host finite element | | t3dElements | [x/y/z] | Direction defining the thickness of 2D quad elements | | reference_header | [at/dat/fe/dfe] | The header that will be used for the output configuration DESCRIPTION: at: the header of the atomic input configuration dat: the header of the atomic input configuration plus maximum FE dispacements fe: the header according the initial FE mesh dfe: the header according the initial FE mesh plus maximum FE displacements |

##4. Elements, nodal positions and displacements

FE2AT needs the information on the elements and nodes of the FE simulation to be stored in three different files. The assignment of the nodes to their elements and the nodal positions can be extracted manually e.g. from the Abaqus input file (file suffix .inp). Comment lines in this file should start with '*'.

To obtain the nodal displacements in the deformed state from an Abaqus output file (.odb), you can use the Python script 'extractDisp.py' in the util/ subdirectory. To extract displacements by execute the script as follows:

abaqus python <FE2ATdir>/util/extractDisp.py <JOBNAME>

Jobname refers to the name of the odb file without the extension (odb)

The nodal displacements are extracted into the files 'FEdispX.dat' where X>0 corresponds to the deformation step.

Updated